Job creation in business services: innovation, demand, polarization(Eng)/ Creación de empleo en servicios empresariales: innovación, demanda, polarización(Ing)/ Enplegu sorkuntza zerbitzu enpresarialetan: berrikuntza, eskaera, polarizazioa(Ing)

IPTS WORKING PAPER on
CORPORATE R&D AND INNOVATION - No. 04/2011
Job creation in business services:

innovation, demand, polarisation
[Francesco Bogliacino, Matteo Lucchese and Mario Pianta]
August 2011
IPTS WORKING PAPER ON CORPORATE R&D AND INNOVATION - 04/2011
JOB CREATION IN BUSINESS SERVICES
The IPTS Working Papers on Corporate R&D and Innovation address economic and policy questions
related to industrial research and innovation and their contribution to European competitiveness. Mainly
aimed at policy analysts and the academic community, these are scientific papers (relevant to and
highlighting possible policy implications) and proper scientific publications which are typically issued when
submitted to peer-reviewed scientific journals. The working papers are useful for communicating our
preliminary research findings to a wide audience, to promote discussion and feedback aimed at further
improvements. They are considered works in progress and are subject to revision.
These IPTS Working Papers on Corporate R&D and Innovation can take the form of more policy-oriented
notes, mainly addressed towards EU policy-makers. They present policy implications derived from our
own research and the views of the most prominent authors in the field, with appropriate references.
This Working Paper (No. 04/2011 – [August 2011]) is issued in the context of the Industrial Research
Monitoring and Analysis (IRMA)1 activities that are jointly carried out by the European Commission's
Joint Research Centre (JRC) – Institute for Prospective Technological Studies (IPTS) and the Directorate
General for Research - Directorate C, European Research Area: Knowledge-based economy.
IRMA activities aim to improve the understanding of industrial R&D and Innovation in the EU and to
identify medium and long-term policy implications. More information, including activities and publications,
is available at: http://iri.jrc.es/ and http://ec.europa.eu/invest-in-research/
The main authors of this paper are Francesco Bogliacino (European Commission, Joint Research
Center- Institute for Perspective Technological Studies, Seville; Rise Group, Medellin), Matteo
Lucchese (University of Urbino) and Mario Pianta (University of Urbino). This paper builds on a
presentation given at the workshop “Business services for innovation, internationalisation and growth”,
held at the University of Rome La Sapienza, 2-3 December 2010. We would like to thank Rinaldo
Evangelista, Valentina Meliciani and the participants for their comments. We also appreciate the
comments and suggestions of three anonymous referees. The analysis conducted in the paper has
been developed in the context of the SID database update, reference to contract 151834-2010A08-IT
by the European Commission JRC.
The IPTS Working Papers on Corporate R&D and Innovation are published under the editorial
responsibility of Fernando Hervás, Pietro Moncada-Paternò-Castello and Andries Brandsma at the
Knowledge for Growth Unit – Economics of Industrial Research and Innovation Action of IPTS/Joint
Research Centre of the European Commission, Michele Cincera of the Solvay Brussels School of
Economics and Management, Université Libre de Bruxelles, and Enrico Santarelli of the University of
Bologna.
Contact information: F. Hervás

European Commission, Joint Research Centre - Institute for Prospective Technological Studies
Edificio Expo
C/ Inca Garcilaso, 3
E-41092 Seville (Spain)
Fax: +34 95 448 83 26; E-mail: jrc-ipts-kfg-secretariat@ec.europa.eu
IPTS website: http://ipts.jrc.ec.europa.eu/; JRC website: http://www.jrc.ec.europa.eu
DG RTD-C website: http://ec.europa.eu/invest-in-research/monitoring/analyses01_en.htm
Legal Notice
Neither the European Commission nor any person acting on behalf of the Commission
is responsible for any use made of this publication.
IPTS WORKING PAPER on CORPORATE R&D AND INNOVATION - No. 04/2011

Full electronic version of the paper can be downloaded at http://iri.jrc.es/
JRC 66487
EUR 24747 EN/4
ISBN 978-92-79-21146-1
ISSN 1831-9408
ISSN 1831-9424
doi: 10.2791/63330
Luxembourg: Publications Office of the European Union
© European Union, 2011
Reproduction is authorised provided the source is acknowledged
Printed in Spain
2
Abstract
The patterns and mechanisms of job creation in business services are investigated in this
article by considering the role of innovation, demand, wages and the composition of
employment by professional groups. A model is developed and an empirical test is carried out
with parallel analyses on a group of selected business services, on other services and on
manufacturing sectors, considering six major European countries over the period 1996-2007.
Within technological activities, a distinction is made between those supporting either
technological competitiveness, or cost competitiveness. Demand variables allow identifying
the special role of intermediate demand.
Job creation in business services appears to be driven by efforts to expand technological
competitiveness and by the fast growing intermediate demand coming from other industries;
conversely, process innovation leads to job losses and wage growth has a negative effect that
is lower than in other industries. Business services show an increasingly polarised
employment structure.
JEL Classification: J20, J23, O30, O33
Keywords: Business Services, Innovation, Employment
3
1 Introduction
In recent decades a number of service industries closely linked to business activities in the
rest of the economy – business services – have shown a strong dynamism in terms of
innovation and employment growth; they create and diffuse knowledge, are extensive users of
ICTs and have a strong impact on the activity of manufacturing and other services; in this way
they are emerging as potential drivers of economic growth and job creation.
A large literature has recently emerged on the importance and potential of business services
in advanced economies, emphasising the outsourcing of activities from manufacturing and the
rise of knowledge-based activities with a potential for spreading innovations across all the
economy (see, among others, Kox, 2001; Miles, 2007).
From a European policy perspective, the growing interconnections between business services
and manufacturing require a deep analysis of dynamics of these industries, in order to achieve
a better calibration of industrial policies.1
In this article we identify the subset of service industries that share the above characteristics
and contrast them with other services and manufacturing sectors, focusing on six major
European countries - Germany, France, Italy, The Netherlands, Spain and the UK - over the
period 1996-2007. After considering the patterns of growth for value added, employment,
productivity, innovative activities and qualification of jobs in terms of professional categories,
we identify the following industries as business services: Post and Telecommunications
(NACE Division 64), Renting of Machinery and Equipment (NACE Division 71), Computer and
Related Activities (NACE Division 72), Research and Development (NACE Division 73), and
Other Business Activities (NACE Division 74).
While some heterogeneity exists within this group, they systematically outperform other
services and manufacturing sectors in the aspects listed above (as will be shown in section 3)
and share a key role as providers of high-value inputs to the rest of the economy (on this role,
see Evangelista, Lucchese and Meliciani, 2011). In the six countries we investigate, this group
of business services accounts for 14% of value added and 15% of employees in 2007, but
their employment growth has been much faster than that for the whole economy.
Financial services, banking and insurance are omitted from this definition because their
growth has been highly dependent - especially in some countries - on speculative bubbles that
have led to the 2008 crises and job creation has been illusory. Moreover, their innovative
stance has been more limited and restructuring practices were significant well before the crisis
4
exploded. We consider these industries in the aggregate of ‘other services’ as a term of

comparison alongside manufacturing, for the performances of business services.
The key question we address in this article is why have business services been able to
achieve such results in terms of job creation? What are the mechanisms that have supported
such performances? And how do they differ from the mechanisms operating in other services
and manufacturing?
The literature on structural change provides us with a strong perspective for investigating such
patterns. We can expect technological change on the supply side and its growth on the
demand side to be two major forces driving the business services’ growth (Pasinetti, 1993).
We develop a model and carry out empirical tests that build on this approach, but also
consider important novelties from different streams of research. First, from innovation studies,
we adopt the distinction between innovative efforts aimed either at technological
competitiveness, based on innovation in products and markets, or at cost competitiveness,
relying on supplier-driven process innovations (Pianta, 2001). These strategies represent
different technological trajectories that help explain the evolution of different industry groups.
A large literature has shown that these strategies have contrasting effects on employment at
the industry level, with the former supporting job creation and the latter leading to job losses
(Pianta, 2005; Mastrostefano and Pianta, 2009; Bogliacino and Pianta, 2010).
Second, the demand side is considered; the relationship between demand, productivity and
employment has mainly been studied for manufacturing industries and we are careful to
consider both intermediate and final demand, and both domestic and foreign sources, in order
to identify the key factors supporting employment growth. We have taken these data from
Input-Output tables and especially examine business services’ role in providing advanced
inputs to other industries.
In fact, intermediate demand has been the fastest growing element of demand in European
countries, reflecting an economic structure that has become less vertically integrated and
more interdependent across sectors; business services have played a key role in this process.
Moreover, the trade of intermediate goods has contributed to higher internationalisation,
leading to production outsourcing and offshoring. In some service sectors internationalisation
is low, foreign demand is modest and domestic markets remain crucial; the result has often
been a low degree of competition and lower productivity growth. Our analysis will clarify the
position of business services within this complex dynamics.
1
The relationship between manufacturing and associated services is quoted in the Industrial Policy
flagship of Europe 2020 strategy, such as the role played by ICT and technology along the value
chain.
5
Third, from labour market studies we consider the relevance of wage growth in order to test
whether job creation in business services is affected by the ‘neoclassical’ trade off between
wage and employment growth.
Fourth, we control for the market structure, using both a proxy of concentration (Herfindahl
index) and the average firm size in the sector, in order to account for both competitive
pressure and industry evolution dynamics (entry patterns, etc.).
Finally, given that recent literature has emphasised the quality, and not just the quantity, of
employment, we consider the way employment is composed in terms of professional groups
as important and examine events that have occurred in business services, comparing them to
other services and manufacturing. In particular, we used data on the shares of four major
professional groups – managers and professionals; clerks; craft workers; manual workers -
and investigate the occupational structure’s growing polarisation, with net job creation mainly
in the top and bottom groups (Nascia and Pianta, 2008).
Different models are proposed in order to explain the driving forces of employment change in
business services and other industry groups, highlighting mechanisms linked to technological
strategies, demand dynamics and labour markets.
Empirical work and econometric tests are based on the database developed at the University
of Urbino, the Sectoral Innovation Database (SID) (Lucchese and Pianta, 2011) which
provides a broad description of different dimensions of industry activities. It integrates
innovation data from EUROSTAT Community Innovation Surveys (CIS) with a large number of
indicators on economic and demand variables from the OECD STAN database and the OECD
Input-Output Tables. The SID dataset covers 22 manufacturing sectors and 17 service sectors
- NACE REV.1 subsections - for six European countries - Germany, France, Italy, The
Netherlands, Spain and the UK, which represent around 80% of the EU-12’s value added
during the two decades under consideration.
The paper is organised as follows: Section 2 reviews the existing literature on business
services and employment; Section 3 offers a descriptive analysis on the dynamics of business
services and their employment structure, considering the role of innovation and demand;
Section 4 presents the model and the econometric strategy; Section 5 shows the results; and
Section 6 provides a conclusion.
6
2 Literature review
2.1. The growth of business services
In the last decades, the weight of business services in the economy has rapidly grown. Many
studies have analysed the determinants, with different levels of analysis pointing out the
diversified patterns across time and countries.
Kox (2001), Kox and Rubalcaba (2007) and Miles (2007) offer an exhaustive description of
channels through which business services have risen in the last decades both in terms of
value added and employment growth. Their rise is related to activity outsourcing, which were
previously carried out within manufacturing firms. In this case, job creation and value-added
growth in the business services industry have replaced the drop in manufacturing production.
As business services emerge as independent activities, their innovative potential becomes

stronger, supported by a high level of interaction between upstream and downstream firms:
business services supply ‘intangible assets’ that contribute to specialising productive
processes. The use of these inputs imply a “different way of organising social production,
allowing a better spread of the advantages of knowledge specialisation, more external scale
economies, and a higher-level growth path”.2 In this way, the nature of the intermediate
relationships between firms seems to have changed, supporting the possible expansion of
value added.
Through the use of Input-Output Tables, some exercises of decomposition have documented
the importance of the role of other sectors’ intermediate demand in business-service growth.
Other studies have emphasised the rise of final demand, showing a further evolution of
relationships between services and end users. In the 1990s, the growth of business services
was also related to deregulation and privatisation of economic activities, which led to the rise
of private specialised suppliers, replacing activities formerly carried out by the public sector
(Kox, 2001; Pilat and Wölfl, 2005; Savona and Lorentz, 2006).
In general, the strong growth of employment in business services has often been considered a
by-product of the scarce productivity performances of these sectors. ‘Baumol’s disease’
(Baumol, 1957) argues that the shift in employment from manufacturing to services is due to
the structural productivity differential between the two sectors. While some studies have
2
Kox and Rubalcaba (2007).
7
shown that these dynamics are dependent on country and sectoral conditions (Baker, 2007),
‘Baumol’s disease’ is involved by an increasing weight of services which could reduce overall
economic system growth.
The specificity of business services – which often have higher productivity performances than
‘traditional' services – has been pointed out. A major factor here is the importance of
knowledge and innovation in business services, often related to the application and diffusion
of ICTs; the provision of business services and the vast knowledge externalities that originate
from the sector may contribute to better performances in the whole economy. These aspects
and the diffusion of knowledge through the relationships that are established among industries
have been documented extensively in the literature (Antonelli, 1999; Camacho and Rodriguez,
2007; Evangelista et al., 2011 in this special issue).
2.2. Innovation, demand, professions and employment
The literature reviewed above explains employment growth in business services as a result of
the structural change in advanced economies. However, the importance of knowledge and
innovation as a source of job creation in this emerging industry requires consideration for the
role that innovation plays in affecting employment. Deeply discussed since the Classical
Economics era, the relationship between technological change and employment has
traditionally been empirically investigated only considering the manufacturing industry, starting
with the pioneering works of Freeman, Clark, and Soete (1982) and Freeman and Soete,
(1987, 1994).3
Studies focusing on the firm level have generally found a positive relationship between
innovation and job creation (see for instance Van Reenen, 1997; Piva and Vivarelli, 2005;
Piva, Santarelli and Vivarelli, 2005; Evangelista and Savona, 2003; Mansury and Love, 2008).
Greenan and Guellec (2000) however find that the positive employment impact of product and
process innovation at the firm level disappears at industry level (where only new products lead
to new jobs).
In fact, innovative firms face no demand constraint and when they are more efficient – through
either new products or processes – they can expand output and jobs also at the expense of
competitors. Conversely, at the industry level the overall potential for job creation is
constrained by increasing industry demand and by the dynamics of labour productivity.
Building on the Schumpeterian distinction between product and process innovation, Pianta
3
Surveys of current literature - with different perspectives and coverage - are in Addison and Teixeira,
2001; Acemoglu, 2002; Chennells and Van Reenen, 2002; Spiezia and Vivarelli, 2002; and Pianta,
8
(2001) has proposed a distinction between the strategies of technological or cost

competitiveness and has suggested the need to consider demand factors for explaining
employment outcomes. Such a distinction identifies the predominant orientation of sectors in
terms of the nature of the innovative efforts produced. The strategy of technological
competitiveness is associated with a general tendency to internal innovative activity, a
prevalence of product innovations and a propensity to search for new markets; the cost
competitiveness strategy is related to a prevalence of concerns about cost efficiency and
labour saving process innovations. This distinction is based on the idea that the economic
sectors are characterised by a different technological trajectory that shapes the perspective of
sectors’ growth. In fact, it is the unfolding of these trajectories that brings about a variety in
growth performance of sectors and, as a consequence, a continuous change in economies’
internal structure. This approach has been integrated with the consideration of demand
factors, in order to account for the need of ‘demand-pull’ effects for creating the conditions for
achieving innovations’ potential and supporting job creation.
Following this approach, the technology-employment link has been investigated at the industry
level by studies that have extensively used evidence from innovation surveys. Vivarelli,
Evangelista and Pianta (1996) and Pianta (2000, 2001) examined the weak European job
performance in the 1990s, showing it to be related to low levels of product innovation, stable
wages and low-demand dynamics. Similar results are found in Antonucci and Pianta (2002)
and Evangelista and Savona (2003); the latter study focuses on the employment patterns in
service industries in Italy, where job creation occurs mainly in small, technology-driven firms.
Mastrostefano and Pianta (2009) examine the effects on employment that result from different
types of innovation, labour market factors, and demand dynamics. When the analysis of the
impact of innovation on job creation is studied in the long-term, the ‘neoclassical’ negative
relationship between wage and job growth seems to be less relevant, while the
‘Schumpeterian’ job creating effect on innovation’s market impact emerges as a major factor.
This research has been extended to services in Bogliacino and Pianta’s study (2010), where
the relationship between technological regimes and patterns of employment growth is studied
through the introduction of a Revised Pavitt Taxonomy. The latter is able to describe the
different technological opportunities of manufacturing as well as service sectors. Again, the
results show that product innovation has a positive impact on employment in Science-Based
and Specialised Suppliers sectors – both in manufacturing and in services – while labour-
saving effects prevail in the case of sectors more oriented towards process innovation (Scale
and Information Intensive and Suppliers Dominated sectors). In identifying the mechanisms of
job creation, the general distinction between manufacturing and services appears less
2005.
9
relevant than the specificity of each Revised Pavitt class.

Other studies have focused on the role of demand in supporting technological change and
performances, showing its crucial role in creating the conditions for growth (Crespi and Pianta,
2008; Bogliacino and Pianta, 2008 and 2011).
Finally, a growing literature has addressed changes in the relative composition of employment
by professional skills (Acemoglu, 2002). The dominant interpretation is that the emergence of
new technologies has led to a pattern of skill-biased technological change as innovations
replace unskilled labour with workers with higher competences, which are complementary to
the new technologies. Job opportunities for blue collar workers in the labour market worsen
and the resulting inequality is presented as a ‘natural’ effect of technological change.
More recent studies focused on the ability of computers to replace routine-task workers, while
activities such as decision making (by managers) and menial jobs (such as cleaning, by the
least-skilled workers) cannot be automated. The outcome is a polarised employment structure
where the share of middle skills is falling (Autor et al., 2003; Autor et al., 2006; Autor and Dorn,
2010; Moose and Manning, 2007).
A more detailed investigation has used data on employees by professional qualifications in 36
manufacturing and service industries for five EU countries, considering four professional
groups: Managers, Clerks, Craft workers and Manual workers (Nascia and Pianta, 2008).
Distinct patterns emerge when industries are grouped based on their patterns of technological
change, i.e. technological competitiveness in high innovation industries and cost
competitiveness in traditional sectors. The overall skill intensity is substantially higher in the
former industry group, and from 2000 to 2003 a clear employment polarisation pattern
emerged, with more jobs being created for managers (+2% a year) and manual workers
(+1.2%) and job losses for clerks (-0.2%) and skilled workers (-2%). Rather than a linear shift
from low-skilled jobs to high-skilled, white collar employment, as predicted by the skill-biased
hypothesis, a clear polarisation pattern is found. When the determinants of employment
changes are explored separately in the four professional groups, different relationships
emerge. Product innovation and higher education lead to more jobs for high-skill categories;
process innovation and cost-reduction strategies destroy jobs for craft workers (ibid.).
Four main lessons can be drawn from such different streams of literature. First, the industry
level is particularly appropriate to investigate the innovation-employment link as it is able to
take into account the overall patterns of structural change and the demand constraint that
operates for industries. Second, there is a need to break down technological change into
different strategies – technological vs. cost competitiveness – which have contrasting effects
on employment. Third, the very relationships defined by models need be adapted to the
specificities of industry groups characterised either by different technological trajectories or by
10
the particular nature of business services. Fourth, the quality of jobs has to be taken on board
in the analysis, with attention to the professions and skills present in jobs created and lost.
The approach we will follow herein will be based on these lessons from recent literature..
3 Empirical evidence
Job creation in services is not uniform. If we examine employment growth from 1996 to 2007
in the six major European economies (Germany, France, UK, Italy, Spain and The
Netherlands) a substantial divide emerges between traditional and business-service activities.
In Graph 3.1., services are grouped by Trade and Leisure (NACE Divisions 50-55), Transport
and Storage (NACE Divisions 60-63), Finance, Insurance and Real Estate (NACE Divisions
65-70) and Business Services (NACE Divisions 64, 71-74): the latter grow twice as much as
other services do, while the rising weight of finance in the economy is not matched by a rapid
growth in employment. The variability within macro sectors is however not negligible: in
business services, Post and Telecommunications have experienced void employment
dynamics, while Computer and Related Activities grew 7% per year on average among the
countries considered. Other Business Activities (NACE Division 74), which represents more
than half of employment of all business services, grew 5%.
This pattern looks stable among the countries: business services show higher dynamics than
overall services; the gap is marked for Germany, where their growth is twice as much as the
growth in the whole service sector. The higher services’ growth rates as a whole in Spain and
Italy are associated to the late development of the tertiarisation process. A more detailed
picture for each service industry is provided in the Appendix.
Graph 3.1. Compound annual rates of employment growth from 1996 to 2007. Averages
among countries and sectors.
11
Source: SID database
Graph 3.2. explores the employment dynamics from 1996 to 2007, considering the trend of
the manufacturing sector, overall services and business services. Their dynamics appear to
be similar, although the performance of manufacturing is strickingly lower than that of
services: from 2000 to 2007, the European manufacturing system has destroyed a large
quantity of jobs, replaced by a recovery in service and (especially) business services
industries. The latter have strongly suffered the economic crisis from 2000 to 2003 and have
not returned to rapid growth until 2007.
Graph 3.2. Annual rates of employment growth. Average values among countries.
Looking at Graph 3.2., three phases of development can be identified: from 1996 to 2000
there is a strong employment growth in services, pushed by job creation in business services;
from 2000 to 2003 the destruction of employment in manufacturing is not counterbalanced by
a growth of employment in services; from 2003 to 2007 manufacturing reduces its job losses
and a strong divergence emerges in the formation of employment between traditional and
business services. The dynamics of business services seem to be more subject to the
economic cycle than other services, as Kox and Rubalcaba (2007) have suggested: this
12
seems to be due to the high integration of business services with the manufacturing sector
and to a high level of labour flexibility.
The different growth of services brings about a change in the qualitative development of jobs
offered.4 Many studies have in fact focused on a change in the composition of skills required
following the structural change process.
Look at the composition of employment by professions (Graph 3.3.) in 2000. Manufacturing
and services are characterised by a different composition: services are marked by a higher
share of Managers and Clerks while Craft and Manual workers constitute the greater part of
manufacturing employment. Educational levels distribution is obviously related to differences
in professions. A general process of up-skilling is present in all sectors, although the
differences between manufacturing and services seem to be less relevant. Additional
evidence on the employment change rates in each of the four professional groups is provided
in the Appendix.
Graph 3.3. Composition of employment in 2000 by professions and by education.

Averages across countries.
4
Due to the availability of data for professions and educations, five countries are considered here (The
Netherlands are excluded) for a reduced period of time, from 2000 to 2003. Professions are grouped by
Managers and professionals, Clerks, Craft Workers and Manual Workers (based on the International
Standard Classification of Occupations ISCO88 COM nomenclature at one digit level). Educational
levels are classified by university, secondary and primary education (according to the ISCED
nomenclature at one digit level).
13
Besides these patterns concerning the quantity and quality of employment, it is also important
to consider how demand evolves. Business services are strongly dependent on intermediate
demand, while other services rely on final consumption and manufacturing industry shows
equal share of intermediate demand and exports. The evidence is provided in Graph A3 in the
Appendix. These different demand patterns have important consequences on our explanation
of employment growth determinants in business services as opposed to the rest of the
economy.
4 The model and econometric strategy

This section explores the factors contributing to job creation in Europe, linking supply and
demand factors. The potential for job growth can be reduced by the occurrence of restrictive
conditions on demand which sharpen the results of the technological competition among firms
and sectors. Through the information drawn from the Input/Output Tables from OECD, we can
break down the different demand components and study their impact on employment growth.
The innovative activity of sectors is investigated by considering the dominance of a strategy
based on technological and cost competitiveness (Pianta, 2001). These strategies are
supposed to have contrasting effects on employment. On the one hand, technological
competitiveness, rooted in quality advantages and the introduction of new products, opens up
new opportunities for demand and employment growth, although this effect can be
counterbalanced by the formation of monopoly rents that can reduce the creation of value
added. On the other hand, cost competitiveness, based on process innovation, engenders
productivity improvements which are largely due to job losses. A positive effect on
employment can also result from reduced prices that can stimulate new demand. However,
the prevailing of the first strategy over the latter is generally supposed to provide higher job
creation opportunities.
Employment growth is also dependent on the dynamics of demand. When its composition is
considered (household and intermediate consumption, investments, and exports), different
competitive regimes are acknowledged. The manufacturing sector is associated with a strict
international competition which asks for a continuous search for higher levels of productivity in
order to sustain exports growth. Service sectors are related to internal demand and to the
process of rising fragmentation of production where competitive pressure is obviously
14
reduced.
Interestingly, the Sectoral Innovation Database links all of these factors in a single framework.
4.1. Methodology
The following equation is estimated:
empit = α 0 + α1tcit + α 2 ccit + α 3d it + α 4 wit + uit (1)
where emp is the sector’s employment level, tc represents the accumulated knowledge related
with technological competitiveness strategy, cc is cost competitiveness, d is demand and w is
the wage; i and t stands for industry and time, respectively. The model is estimated at industry
level for various countries so the individual observation is a certain industry in a given country.
Our main object is to identify the effect of technology and demand. A typical issue is that the
error term u may be correlated with the regressors, because of time-invariant effects or issues
of simultaneity and/or omitted variables.
If we think of variables in the log scale, we know that by taking the differences we approximate
the rate of change. Through this transformation we eliminate the individual time-invariant
effect (we also control for all time-invariant characteristics, such as country-level institutional
determinants which are likely to play a role) and we can exploit the lag structure to avoid
simultaneity.
Technically, we take long differences (e.g. Caroli and Van Reenen, 2001; Piva et al., 2005),
i.e. differences over a large time span, in order to soften considerably the autoregressive
pattern, which would weaken the validity of the lags. As the log difference approximates the
rate of variation, making the average is just a linear transformation of it (not affecting the main
estimators’ properties), we calculate average rate of change which allows us to merge the
different data sources, which have slightly different time spans.
The model is adjusted for heteroscedasticity (robust estimation) and intra-group correlation at
the industry level, checking for intra-sectoral heterogeneity.
Moreover, weighted regressions are used since industry data are typically grouped data of
unequal size; this provides us with more time-wise stability.
The baseline model becomes:
∆empit = α1∆tcit + α 2 ∆ccit + α 3∆d it + α 4 ∆wit + ∆uit (2)
where the dependent variable is the rate of growth of employment, the variation in the
knowledge stock associated with the technological strategies can be proxied by flows
15
variables coming from innovation surveys, demand growth is measured through the growth
rate of different sources of demand, w is the growth rate of labour compensation per employee
and the last part is the error term, in which the individual time-invariant effects are eliminated.
Following our conceptual framework, we expect the following relationships to emerge:
1) Proxies for technological competitiveness (the share of firms indicating clients as source of
innovation and the share of firms internal source as determinants for innovation) are expected
to have a positive effect on job growth. Employment growth is stimulated by the greater
demand for new products; the positive effect of the new demand is supposed to prevail on the
effects of reduced production due to forming possible monopoly rents in the consumer market
and/or replacing effects for old products;
2) Indicators of cost competitiveness (the share of firms aiming to reduce labour costs and the
share of firms who indicate suppliers as source of innovation) are expected to have direct
labour saving effects. The effect on prices due to increased productivity can however partially
or totally overcome the reduced employment that derives from recourse to a strategy based
on process innovations.
3) Demand variables are expected to have a positive effect on employment; however, while
exports are supposed to drive productivity and employment growth in manufacturing,
employment growth in services is prevalently associated with the growth of intermediate and
household consumption. In sectors where the role of new information technologies is more
intensive and a sustained growth in terms of value added can be observed, economic growth
is principally linked to changes in the production structure that leads to high intermediate
demand from the rest of the economy: 70% of business-service demand only comes from
other sectors.
In terms of control variables, we expect an inverse relationship between wages (labour
compensation growth rate per employee) and employment creation. Labour demand can
increase when labour costs decrease.
We also ran further robustness checks to control for the market structure of industries; we
expect concentration to be negatively related with growth, through a standard lack of
competitive pressure effect. We also add a control for the firm’s average size in the sector: we
expect it to be negatively correlated with growth for the most dynamic industries, where jobs
are disproportionately created by new (and usually very small) firms (Haltiwanger et al., 2010).
4.2. Data
The analysis considers six countries, Germany, France, Italy, The Netherlands, Spain and the
United Kingdom for 21 manufacturing and 17 service sectors.
16
The model is distinctly examined for manufacturing and services sector. In order to
differentiate the dynamics of services, we isolate the effects of business services from other
services. Industries classified as 64 and 71 to 74 by NACE Rev. 1.1 form the first category,
while industries 50, 51, 52, 55, 63, and 65 to 67 are classified as other services. Whenever we
want to separate the effects among groups of industries we run the regressions on the overall
database but allowing for different coefficients. Although the estimates are exactly the same
as those obtained from regressions over the different groups of observations, the estimates
are more efficient if we use the overall sample, given the number of the degrees of freedom.
We believe the time structure deserves a specific comment. We merge three different data
sources (STAN OECD database for employment growth, OECD I-O Tables for demand
variables and CIS data for innovation measures). Having three waves of Input/Output
available (1995, 2000 and 2005) allowed us to consider the demand growth rates for two
periods, 1995-2000 and 2000-2005. In fact, the employment trend described in Section 3
suggests that business cycle impact should remain in the data. We decided not to include the
2001-2003 recession and focus on two periods of employment growth. We excluded CIS3
data and focused on the impact of innovation variables in CIS2 and CIS4 on employment
performance in 1996-2000 and 2003-2007 (in order to consider a lag in the impact of
innovation).
The temporal structure is shown in the following table:
Table 1. The database structure

Type of variable Reference years
Period 1 Period 2
Innovation activities CIS2, 1994-1996 CIS4, 2002-2004
Employment growth (growth rates) 1996-2000 2003-2007
Industries’ demand (growth rates) 1995-2000 2000-2005
Wage 1996-2000 2003-2007
As we can see, data from Innovation surveys always precede data on employment; with
regards to demand, data for the first time started one year before, and data for the second
period finished two years before.
As a further check, the possibility of multicollinearity is checked through the Variance-Inflation-
Factor (VIF) analysis. A preliminary analysis on the distribution of variables has allowed
possible extreme values to be dropped. Moreover, only the significant differences of variables
are considered. The presence of outliers does not affect the values of coefficients in every
model.
The technological proxies extracted from the extensive information in CIS are chosen
17
depending on availability (in order to avoid gaps in the dataset, especially with regards to
service industries) and low number of outliers.
5 Results
5.1. Baseline regressions
First, we ran a general regression with the baseline model.
Table 2. The general model

Source: SID. Dependent variable: rate of change in total employment. Standard errors are
robust and clustered by industry, reported in brackets. * is significant at 10% level, ** at 5% and
***
at 1% level. Weights are the number of employees.
WLS
Share of firms that indicate internal source for 0.043
innovation [0.019]**
Share of firms that innovate to reduce labour -0.042
cost [0.017] **
Average rate of growth of labour -0.477
compensation per employee [0.163]***
Export growth rate 0.037
[0.027]
Intermediate demand growth rate 0.202
[0.051]***
Household consumption growth rate 0.074
[0.032]**
Constant 0.433
[0.560]
Observations 349
F-test 9.31
(P value) (0.000)
R2 0.236
As we can observe, the technological and cost competitiveness effect is comparable in

magnitude but with negative sign. An increase by one in the share of firms indicating internal
sources for innovation will add 0.04 percentage points to the average employment growth,
while the opposite will happen if the share of firms that choose labour saving strategy
increases by one point.
Intermediate demand has the highest impact: 1% increase raises by one fifth of percentage
point the annual compound rate of employment. Household consumption had the lowest
impact, while export is not significant.
The wage term has a negative impact, through a standard labour-demand effect.
18
The first differentiation that we want to examine is between manufacturing and services. The
results are in Table 3. If we use similar proxies for technological strategies in both
manufacturing and services we can observe that variables are non-significant for
manufacturing.5 Other proxies that better fit the type of innovation of manufacturing industries,
such as those focusing on the relationship with clients and suppliers in the search process
clearly make the technological-cost competitiveness distinction emerge.
Table 3. Manufacturing versus services

***
at 1% level. Regression was ran on the overall sample with coefficient differentiated
between manufacturing and services. Weights are the number of employees.
Manufacturing Services
Share of firms that indicate 0.033
clients as source of innovation [0.015]**
Share of firms that indicate -0.052
suppliers as source of [0.021]**
innovation
Share of firms that indicate 0.082
internal source for innovation [0.025]*
Share of firms that innovate to -0.058
reduce labour cost [0.35]**
Average growth of labour -0.600 -0.416
compensation per employee [0.142]*** [0.158]*
Export growth rate 0.049 -0.004
[0.039] [0.031]
Intermediate demand growth 0.016 0.232
rate [0.042] [0.068]***
Household consumption growth 0.026 0.093
rate [0.019] [0.057]*
Constant 0.316
[0.377]
Observations 349
F-test 9.31
(P value) (0.000)
R2 0.236
When different variables are introduced for manufacturing and services, innovation strategies
are qualitatively different but quantitatively comparable in their effects through the two
groupings. Indeed, technological and cost competitiveness are well captured respectively
through the share of firms indicating clients and those indicating suppliers as source of
innovation. A t-test of the difference of impact between the manufacturing and services proxy
5
The results are available from the author upon request.
19
does not reject the hypothesis of a zero difference at 5% confidence level. This is a main
result of the analysis conducted using SID data (Bogliacino and Pianta, 2010 and 2011); its
validity is confirmed in this paper.
Adding 1% to the share of firms that choose technological competitiveness translates into a
higher employment growth rate of +0.03% for manufacturing and +0.08% for services.
Increasing the share of firms who opt for cost competitiveness by 1%, subtracts around 0.05-
0.06 percentage points from the employment growth rate.
The demand effect is largely confined to services, since for manufacturing they are never
significant (meaning that the effect is captured by productivity): again the largest impact
comes from intermediate demand, where 1% increase in demand adds 0.23% to the average
employment growth rate.
5.2. Capturing the peculiarity of business services

In the following Table we extend the baseline model to account for the heterogeneity between
manufacturing, business services and the rest of the service sector.
Table 4. The role of business services

***
between manufacturing, business services and other services. Weights are the number of
employees.
Manufacturing Business services Other services
Share of firms that 0.029
indicate clients as source [0.015]*
of innovation
Share of firms that -0.054
indicate suppliers as [0.021]**
source of innovation
Share of firms that 0.068 0.034
indicate internal source [0.036]* [0.024]
for innovation
Share of firms that -0.040 -0.059
innovate to reduce [0.065] [0.032]*
labour cost
Average rate of growth -0.598 -0.561 -0.601
of labour compensation [0.146]*** [0.163]* [0.114]***
per employee
Export growth rate 0.047 -0.068 -0.046
[0.040] [0.052] [0.023]*
Intermediate demand 0.003 0.245 0.270
growth rate [0.043] [0.122]** [0.059]***
20
Household consumption 0.027 0.043 0.152

growth rate [0.020] [0.066] [0.064]**
Constant 0.583
(0.403)
Observations 347
F-test 13.06
(P value) (0.000)
R2 0.510
As we can observe, there are three sorts of trajectories emerging: manufacturing shows both
technological and cost competitiveness, but no demand effect; business services are focused
on technological competitiveness and intermediate demand is its main source of demand; and
finally, other services are cost competitiveness intensive and intermediate and household
demand are the main job creation drivers.
On the orders of magnitude, we can observe that rising by 1% the share of firms that choose
technological competitiveness three basis point are added to the employment growth in
manufacturing, and seven basis points in business services. If we were to increase the share
of firms that choose cost competitiveness by 1%, around 0.05% is subtracted from the
average employment growth in manufacturing and other services. Demand counteracts this
direct labour saving effect in the rest of the service sector, but not in manufacturing, where this
compensatory mechanism does not work, at least in the time span considered.
It is also very important to stress that this model captures the heterogeneity quite well: in fact,
it is able to account for around 50% of total variance.
5.3 Controlling for market structure

In order to observe the robustness of the above specification, we used two variables that
capture market structure: one was the Herfindahl6 index calculated at a two-digit level; the
second one is the average firm size.
While average firm size can be drawn from CIS, the Herfindahl index is taken from the EU-
KLEMS database (see O'Mahoni et al., 2008). For some sectors, Herfindahl estimates are
provided at a three-digit level. In order to obtain data at two digits, we weighted data at three
digits with the respective share of production for the sector total.
To avoid potential multicollinearity problems, we eliminated the non-significant demand
variables from the regressions. We maintained intermediate demand given that we did not
wish to completely eliminate demand from the manufacturing equation.
6
Herfindahl Index is a concentration index calculated as the sum of squares of the shares: H = ∑ si2
i
21
The concentration shows a negative and significant effect in manufacturing industries,

indicating that rent consolidation negatively affects job creation, because competitive pressure
to invest is lacking. No significant effect is found for services.
On the contrary, average firm size is showing a negative and significant impact for business
services. The latter are leading the structural change process, thus the job creation effect is
mainly driven by entry of new small firms, coherently with its dynamism. No significant effect is
found in the other two subgroups.7
We summarise the effect by running the regression using the Herfindahl for manufacturing
and the average firm size for services.
Table 5. The baseline model with market structure determinants

***
employees.
Share of firms that 0.038
indicate clients as source [0.015]**
of innovation
Share of firms that -0.048
indicate suppliers as [0.021]**
source of innovation
Share of firms that 0.078 0.013
indicate internal source [0.033]** [0.029]
for innovation
Share of firms that 0.025 -0.035
innovate to reduce [0.062] [0.032]
labour cost
Herfindahl Index -2.493
[1.400]*
Average firm size -0.821 -0.889
[0.343]*** [1.152]
of labour compensation [0.139]*** [0.152]*** [0.155]***
per employee
Export growth rate 0.144
[0.023]*
Intermediate demand 0.059 0.270 0.280
growth rate [0.043] [0.112]** [0.058]***
Household consumption 0.169
growth rate [0.067]**
Constant 0.465
7
These results are available from the authors upon request.
22
(0.470)
Observations 302
F-test 11.13
(P value) (0.000)
R2 0.524
In summary, employment in manufacturing is largely explained by the positive effect of

technological competitiveness and the negative direct effect of cost competitiveness; it is also
negatively affected by concentration. In traditional and finance services, dynamics are largely
explained by demand evolution (and at least for the latter part, also partly artificially inflated by
bubble processes).
Finally business services show a Schumpeter Mark I trajectory, where entry by new firms and
product innovation are the key determinants of the sector’s evolution, coupled with
intermediate demand’s significantly important role, related with the process of de-
verticalisation and globalisation.
5.4 Controlling for occupational structure

An interesting insight that we can add using SID is the relationship between the employment
structure and polarisation. At this stage we can only discuss association and not the causality
chain, for two main reasons: a) on the one hand, we are missing information due to limited
data, the sample is significantly reduced and we were confronted by a major problem related
to using the lag structure to identify effects; b) on the other hand, it is really problematic to
identify causality chains; both phenomena are influenced by technology (Acemoglu, 2002) and
quantity and quality of labour are interrelated.
Data on Professions composition of sectors include measures drawn from the national Labour
Forces Surveys (LFS). It allows us to collect sectoral data on professions, education of
employees within each sector for five European countries (NACE Rev. 1) (The Netherlands is
excluded) through harmonised surveys that do not have confidentiality problems. Data were
collected from 2000 to 2003. The former is associated with the first period and the latter with
the second one in our SID database. As we discussed above (see footnote 4), we divided the
10 International Standard Classification of Occupation (ISCO) occupational classes into
managers, clerks, crafts and manual workers.8
In Table 6, we ran the regression using a polarisation measure, computed as the shares of
managers and manual workers in total employment over the total number of shares of clerks
8
National Labour Force Surveys (LFS) are the main sources of data on professions in Europe, as they
provide comparable and detailed information on professions categorised by of employing firms’
economic activity sector (see Lucchese and Pianta, 2011).
23
and crafts workers.

We added controls for demand and market structure. We eliminated the technological
variables because, as discussed above, employment composition tends to capture the same
dynamics as knowledge accumulation. As we stressed, this is a measure of association and
not of causation, so we are not particularly concerned about omitting variable problems.
Table 6. Employment growth and polarisation

Source: SID. Dependent variable: Rate of change in total employment. Standard errors are
***
employees.
Polarisation -0.935 0.690 -0.318
[0.204]*** [0.181]** [0.150]**
Herfindahl Index -3.110
[1.380]***
Average firm size 1.279 -0.097
[0.112] [1.279]
of labour compensation [0.169]*** [0.419]*** [0.155]***
per employee
Export growth rate -0.056
[0.154]
Intermediate demand 0.095 -0.145 0.179
* **
growth rate [0.047] [1.136] [0.078]**
Rate of growth of 0.138
household consumption [0.115]**
Constant 0.465
(0.470)
Observations 148
F-test 14.74
(P value) (0.000)
R2 0.616
Coherently with the descriptive evidence presented in Section 3, the results show a
fundamental difference between manufacturing and business services: in the former the
employment structure is associated with the more ‘fordist’ system in which employment growth
is associated with middle-range human capital. Business services behave in the reverse
manner: their growth is concentrated into a very high and a very low (mainly those constrained
not to be outsourced) position in the job ranking and as such, polarisation is strongly
associated with structural change towards those industries (for an explanation of the
technological determinants see Autor and Dorn, 2010 and Nascia and Pianta, 2008).
Other services show a pattern similar to manufacturing but with a significantly lower
24
association, probably due to high heterogeneity inside the group.
6 Conclusions and policy implications

Building on the well-documented differences between manufacturing, business services and
other services in terms of structural change, we have shown that business services have
outperformed job creation in other economic activities in major European countries over the
last two decades. Other services show heterogeneous patterns, ranging from employment
growth in selected niches to job losses due to restructuring in retail trade, banking and
financial activities. Conversely, employment in manufacturing has shown a general downward
trend, with few exceptions in high-innovation sectors and in the industries of greater national
specialisation among European countries.
We have investigated the fundamental mechanisms leading to such patterns of structural
change, paying attention both to technological change on the supply side, and to the sources
of demand. Analysing the role of technological change has confirmed the presence of two
contrasting effects. First, industries where a technological competitiveness strategy prevails –
based on knowledge creation, product innovation and development of new markets – show
strong job-creation ability. Second, industries dominated by a search for cost competitiveness
– relying on acquisition of technologies from suppliers, labour saving strategies and process
innovations – tend to use technologies to replace labour and show serious job losses. These
findings are consistent with the extensive literature reviewed in Section 2, which we have
systematically compared first between manufacturing and services overall, and then
considering business services and other services independently. In all cases we have found
that these relationships are maintained. The nature of the impact of technological change on
employment is the same in all industries, with innovations focusing on new products that are
capable of increasing employment, while new processes lead to job losses. The pace at which
this happens, however, is different in business services and in the rest of the economy. The
job creation mechanisms in the former indeed appear to be stronger than in the latter, and
some of the limits to job creation (e.g. the negative effect of high wage growth) are weaker in
business services than in other industries.
In fact, we have found that different variables capture these contrasting patterns in
manufacturing and service industries in a more effective way; in the former the search for
technological competitiveness is better documented by the orientation towards clients, while in
the latter by the use of internal sources as main source of innovation.
25
Conversely, a cost-competitiveness strategy is reflected by the suppliers in the manufacturing

sector and by aiming to reduce labour costs in the services sector.
Structural change, however, is not the result of developments in the supply side alone. On the
demand side industries are affected by different sources of intermediate and final demand –
both domestic and foreign – that grow at differing paces and shape the expansion of
industries’ economic activities and jobs (Pasinetti, 1981).
The strongest differentiation between manufacturing, business services and other services
emerges on the demand side. In explaining industries’ employment growth, we have shown
that manufacturing job decline has been affected by the lack of demand.
Conversely, jobs in business services grow as a result of the strong dynamics of intermediate
demand, as all industries increasingly need the research, software, consulting, accounting,
communication services offered by business services. This is shaping a new pattern of inter-
industry interdependencies where business services play a key role in the whole economic
system’s competitiveness (Evangelista, Lucchese, Meliciani, 2011).
Other services play a role in different demand sources: by far, demand is the main job growth
driver, contrasted with a technological dimension which is mainly labour saving, being focused
on cost competitiveness.
Job creation mechanisms in business services therefore appear distinct in relation to the rest
of the economy: new knowledge and new products, fast-growing demand from other industries
and greater space for wage increases characterise the business service employment growth.
When we control for market structure we see that concentration is negatively associated with
job growth in manufacturing, while in business services there is a clear negative relationship
between the average firm size and job growth, suggesting that the industry growth is driven by
entry of new firms and growth of small ones.
Finally, when we look at the dynamics of knowledge generation and accumulation through the
structure of human capital, we observe that there is a clear distinction between manufacturing
and services. Using four occupational categories (share of managers, clerks, crafts and
manual workers) we built a polarisation measure as the ratio of the share of top and bottom
occupation (manual plus managers) over the sum of shares of crafts and clerks. The resulting
variable is positively associated with job growth in business services and negatively in
manufacturing and other services. These effects capture the different skill composition
required by the technology set available in manufacturing (more ‘Fordist’) and in the new
business services, where ICTs are prominent.
There are a few policy implications to stress:
26
a) Since intermediate demand is becoming more and more important due to globalisation
processes, demand management policies are very relevant for the process of job creation.
Negative shocks have stronger effects in the economy;
b) Given the entry of new firms plays an important role in the dynamics of business service
employment growth, it suggests that entry barriers should be eliminated in the most dynamic
parts of the economy;
c) Where cost competitiveness is stronger, policies facing labour reallocation are fundamental
to soften the impact of restructuring;
d) Business service growth induces a transformation of the skill mix required by the economy,
i.e. increasing polarisation and reducing middle-class jobs, which requires proper intervention
of the authorities to prevent negative effects of inequality.
Acknowledgements
This paper builds on a presentation to the workshop “Business services for innovation,
internationalisation and growth”, held at the University of Rome La Sapienza, 2-3 December
2010. We would like to thank Rinaldo Evangelista, Valentina Meliciani and the participants for
their comments.
We acknowledge financial support from IPTS with the Project 151834 - Gathering quantitative
evidence of the impact of corporate R&D and innovation on employment: An analysis of
demand, supply and institutional factors at industry and firm level.
References
Acemoglu, D. ‘Technical change, inequality and the labour market’, Journal of Economic
Literature, 40, 1, 2002, pp. 7-72.
Addison, J. and Teixeira, P.Technology, employment and wages, Labour, 15, 2, 2001, pp.
191-219.
Antonelli, C. The Microdynamics of Technological Change, Routledge, London, 1999.
Antonucci T., Pianta M. ‘The employment effects of product and process innovations in
Europe’, International Review of Applied Economics, Vol. 16, 2002, pp. 295-308.
Autor, D., Levy, F., Mundane, R. ‘The Skill Content of Recent Technological Change: An
Empirical Investigation’, Quarterly Journal of Economics, 118, 2003, pp. 1279-1333.
Autor, D., Katz, L., Kearney, M. ‘The Polarization of the U.S. Labour Market’, NBER
Working Paper, No. 11986, 2006.
27
Autor, D., Dorn, D. ‘The Growth of Low-Skill Service Jobs and the Polarization of the U.S. Labor
Market’. MIT Working Paper, August 2010.
Baker, D. The impact of business-services use on client industries: evidence from input output
data, In: Rubalcaba, L., Kox, H. (eds.), Business Services in European Economic
Growth, Palgrave MacMillan, New York, 2007, pp. 97-115.
Baumol, W.J. ‘Macroeconomics of Unbalanced Growth: the Anatomy of Urban Crisis’,
American Economic Review, Vol. 57, 1967, pp. 415-426.
Bogliacino, F., Pianta, M. ‘The impact of R&D and innovation on economic performance and
employment: a quantitative analysis based on innovation survey data.’ European
Commission-Joint Research Centre, Institute for Prospective Technological Studies
(IPTS-Seville) Report (J03/32/2007), 2008.
Bogliacino, F., Pianta, M. ‘Innovation and employment: a reinvestigation using revised Pavitt
classes’, Res. Policy. 39, 2010, pp. 799-809.
Bogliacino, F., Pianta, M. Engines of growth. Innovation and productivity in industry groups.
Structural Change and Economic Dynamics, 22, 2011, pp. 41-53.
Camacho, J.A., Rodriguez, M. ‘Integration and diffusion of KIS for industry performance’, In:
Rubalcaba, L., Kox, H. (eds.). Business Services in European Economic Growth,
Palgrave MacMillan, New York, 2007(a), pp. 128-143.
Chennels, L., Van Reenen, J. ‘Technical change and the structure of employment and wages:
a survey of the microeconometric evidence’, In: Greenan, N., L'Horty, Y., Mairesse, J.,
(eds.), Productivity, Inequality, and the Digital Economy. A Transatlantic Perspective,
MIT Press, Cambridge, 2002, pp. 175-224.
Crespi, F., Pianta, M. ‘Demand and innovation in productivity growth’, International Review of
Applied Economics, 22, 6, 2008, pp. 655-672.
European Commission 2010. An Integrated Industrial Policy for the Globalisation Era. Putting
Competitiveness and Sustainability at Centre Stage. COM(2010) 614.
Evangelista, R., Lucchese, M., Meliciani, V. Business Services, Innovation and Sectoral
Growth, paper submitted to Struct. Change Econ. Dyn, 2011.
Evangelista, R., Savona, M. Innovation, employment and skills in services. Firm and sectoral
evidence. Struct. Change Econ. Dyn. 14, 2003, pp. 449-474.
Fagerberg, J., Mowery, D., Nelson, R. (eds.). The Oxford Handbook of Innovation, Oxford
University Press, Oxford, 2005.
Freeman, C., Clark, J., Soete, L. Unemployment and Technical Innovation, Pinter, London,
,1982.
Freeman, C., Soete, L. (eds.). Technical Change and Full Employment, Blackwell, Oxford,
1987.
28
Freeman, C., Soete, L. Work for all or mass unemployment? Pinter, London, 1994.
Greenan, N., Guellec, D. ‘Technological innovation and employment reallocation’, Labour, 14,
4, 2000, pp. 547-590.
Haltiwanger, J.C., Jarmin, R.S., Miranda, J. ‘Who Creates Jobs? Small vs. Large vs. Young’,
NBER Working Paper Series, 2010, wp16300.
Kox, H. ‘Sources of structural growth in business services’, CBP Research Memorandum No.
12, the Hague, 2001.
Kox, H., Rubalcaba, L. ‘Analysing the contribution of business services to European economic
growth’, MPRA Paper n. 2003, 2007.
Lucchese, M., Pianta, M. The Sectoral Innovation Database 1994-2009 (SID). Sources and
methodologies, University of Urbino, Faculty of Economics, Discussion Paper, 2011.
Mansury, M.A., Love, J.H. ‘Innovation, productivity and growth in US business services: A firm-
level analysis’ TECHNOVATION, 28, 2008, pp. 52-62.
Mastrostefano V., Pianta M. ‘Technology and jobs’, Economics of Innovation and New
Technology, 18, 7-8: 2009, pp. 729-742.
Miles, I.’Business Services and their users: a literature review’, In: Rubalcaba, L., Kox, H., (eds.),
Business Services in European economic growth, Palgrave MacMillan, New York, 2007.
Moose, M., Manning, A. ‘Lousy and Lovely Jobs: The Rising Polarization of Work in
Britain’, Review of Economics and Statistics, 89, 2007, pp. 118-133.
Nascia L., Pianta M. ‘Skill bias or polarisation? Innovation and job growth in Europe’. Paper
presented at the EAEPE 2008 Annual Conference, Rome, 6-8 November 2008.
O' Mahoni, M., Castaldi, C., Los, B., Bartelsman, E., Maimaiti, Y., Peng, F. .EUKLEMS -
Linked data: Sources and Methods, University of Birmingham,
http://www.euklems.net/data/linked/EUKLEMS_LinkedData_SourcesAndMethods_2207
08.pdf, 2008.
Pasinetti, L. Structural Change and Economic Growth, Cambridge University Press,
Cambridge, 1993.
Pavitt, K. ’Patterns of technical change: towards a taxonomy and a theory’, Research Policy,
13, 1984, pp. 343-74.
Petit, P. Employment and Technological Change, In: Stoneman, P. (Ed.), Handbook of the
Economics of Innovation and Technological Change, Blackwell Publishers, Amsterdam,
1995, pp. 366-408.
Pianta, M. ‘The employment impact of product and process innovations’, In: Vivarelli, M.,
Pianta, M. (eds.), The emploment impact of innovation, Routledge, London, 2000, pp.
77-95.
29
Pianta, M. Innovation, ‘Demand and Employment’, In: Petit P., Soete L. (eds.) Technology and
the future of European Employment, Elgar, Cheltenham, 2001, pp. 142-165.
Pianta, M. ‘Innovation and employment’, In: Fagerberg, J., Mowery, D.C., Nelson R.R. (eds.),
Oxford Handbook of Innovation, Oxford University Press, Oxford, 2005, pp. 568-598.
Pilat, D., Wölfl, A. ‘Measuring the interaction between manufacturing and services’, STI
Working paper 2005/5, OECD, Paris, 2005.
Piva, M., Vivarelli, M. ‘Innovation and employment, evidence from Italian microdata’, Journal of
Economics, 86, 1, 2005, pp. 65-83.
Piva, M., Santarelli, E., Vivarelli, M. ‘The skill bias effect of technological and organisational
change: Evidence and policy implications’, Research Policy, 34, 2005, pp. 141-157.
Rubalcaba, L., Kox, H. (eds.). Business services in European economic growth, Palgrave
MacMillan, New York, 2007.
Savona, M., Lorentz, A. ’Demand and technological determinants of structural change and
tertiarisation - an input-output structural decomposition analysis for four OECD countries’,
LEM Working paper Series, 2005/25, 2006.
Spezia, V., Vivarelli, M. ‘Innovation and employment: a critical survey’, In: , N., L'Horty, Y.,
Mairesse, J. (eds.). Productivity, Inequality, and the Digital Economy. A Transatlantic
Perspective, MIT Press, Cambridge, 2002, pp. 101-131.
Van Reenen, J. ‘Employment and Technological Innovation: Evidence from U.K. Manufacturing
Firms’, Journal of Labor Economics, 15, 1997, pp. 255-84.
Vivarelli, M. The Economics of Technology and Employment: Theory and Empirical Evidence,
Elgar, Aldershot, 1995.
Vivarelli M. Evangelista R., Pianta M., ‘Innovation and employment in the Italian manufacturing
industry’, Research Policy, 25, 1996, pp. 1013-1026.
Vivarelli, M. Pianta, M. (eds.), The Employment Impact of Innovation: Evidence and Policy,
Routledge, London, 2000.
Appendix
Information on employment changes for each service industry and for the all of manufacturing
is provided in Graph A1 below.
Graph A1. Compound annual rates of employment growth. Averages across countries.
30
The analysis of the composition of employment by professions in Graph A2 shows that

business services are characterised by a strong polarisation process that is contrasting with
the expected professional up-skilling process. Employment growth for manual workers in
business services is due to the rise of manual workers in less-qualified and ancillary activities,
together with a strong shift of employment in Post and Telecommunications industry in the
United Kingdom.
Graph A2. Changes in employment by professional groups: 2000-2003.
31
Graph A3 shows the composition of demand for each macro sector. While manufacturing is
characterised by a strong share of exports, traditional and finance services are related to
internal demand. As expected, business services are linked to the demand for intermediate
consumption, while the other components only constitute a modest share of total demand.
Graph A3. Composition of demand in 2000. Averages among countries.
32
European Commission
EUR 24747 EN/4 – Joint Research Centre – Institute for Prospective Technological Studies
IPTS WORKING PAPER on CORPORATE R&D AND INNOVATION - No. 04/2011

Title: Job creation in business services: innovation, demand, polarisation
Author(s): Francesco Bogliacino (European Commission, Joint Research Center- Institute for Perspective
Technological Studies, Seville; Rise Group, Medellin), Matteo Lucchese (University of Urbino) and Mario Pianta
(University of Urbino)
Luxembourg: Publications Office of the European Union
2011
EUR – Scientific and Technical Research series – ISSN 1831-9408
Technical Note – ISSN 1831-9424
ISBN 978-92-79-21146-1
doi:10.2791/63330
Abstract
The patterns and mechanisms of job creation in business services are investigated in this article by
considering the role of innovation, demand, wages and the composition of employment by
professional groups. A model is developed and an empirical test is carried out with parallel analyses
on a group of selected business services, on other services and on manufacturing sectors,
considering six major European countries over the period 1996-2007.
Within technological activities, a distinction is made between those supporting either technological
competitiveness, or cost competitiveness. Demand variables allow identifying the special role of
intermediate demand.
Job creation in business services appears to be driven by efforts to expand technological
competitiveness and by the fast growing intermediate demand coming from other industries;
conversely, process innovation leads to job losses and wage growth has a negative effect that is
lower than in other industries. Business services show an increasingly polarised employment
structure.
The mission of the Joint Research Centre is to provide customer-driven scientific
and technical support for the conception, development, implementation and
monitoring of European Union policies. As a service of the European Commission,
LF-ND-24747-EN-N
the Joint Research Centre functions as a reference centre of science and technology
for the Union. Close to the policy-making process, it serves the common interest of
the Member States, while being independent of special interests, whether private or
national.

Job creation in business services: innovation, demand, polarization(Eng)/ Creación de empleo en servicios empresariales: innovación, demanda, polarización(Ing)/ Enplegu sorkuntza zerbitzu enpresarialetan: berrikuntza, eskaera, polarizazioa(Ing)

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Job creation in business services: innovation, demand, polarization(Eng)/ Creación de empleo en servicios empresariales: innovación, demanda, polarización(Ing)/ Enplegu sorkuntza zerbitzu enpresarialetan: berrikuntza, eskaera, polarizazioa(Ing)

Enviado por

Direitos autorais:

IPTS WORKING PAPER on

CORPORATE R&D AND INNOVATION - No. 04/2011

Job creation in business services:

[Francesco Bogliacino, Matteo Lucchese and Mario Pianta]

Contact information: F. Hervás

IPTS WORKING PAPER on CORPORATE R&D AND INNOVATION - No. 04/2011

JEL Classification: J20, J23, O30, O33

Keywords: Business Services, Innovation, Employment

exploded. We consider these industries in the aggregate of ‘other services’ as a term of

As business services emerge as independent activities, their innovative potential becomes

2.2. Innovation, demand, professions and employment

(2001) has proposed a distinction between the strategies of technological or cost

relevant than the specificity of each Revised Pavitt class.

Source: SID database

Source: SID database

Graph 3.3. Composition of employment in 2000 by professions and by education.

Source: SID database

4 The model and econometric strategy

empit = α 0 + α1tcit + α 2 ccit + α 3d it + α 4 wit + uit (1)

∆empit = α1∆tcit + α 2 ∆ccit + α 3∆d it + α 4 ∆wit + ∆uit (2)

Table 1. The database structure

Table 2. The general model

As we can observe, the technological and cost competitiveness effect is comparable in

Table 3. Manufacturing versus services

5.2. Capturing the peculiarity of business services

Table 4. The role of business services

Household consumption 0.027 0.043 0.152

5.3 Controlling for market structure

The concentration shows a negative and significant effect in manufacturing industries,

Table 5. The baseline model with market structure determinants

In summary, employment in manufacturing is largely explained by the positive effect of

5.4 Controlling for occupational structure

and crafts workers.

Table 6. Employment growth and polarisation

association, probably due to high heterogeneity inside the group.

6 Conclusions and policy implications

Conversely, a cost-competitiveness strategy is reflected by the suppliers in the manufacturing

Source: SID database

The analysis of the composition of employment by professions in Graph A2 shows that

Graph A2. Changes in employment by professional groups: 2000-2003.

Source: SID database

Graph A3. Composition of demand in 2000. Averages among countries.

Source: SID database

IPTS WORKING PAPER on CORPORATE R&D AND INNOVATION - No. 04/2011

Você também pode gostar